Simulating Vehicle Power Consumption, Travel Time, and Routes in Real-Time

The proliferation of sensors and data allows traffic simulators to assist drivers by predicting their route, travel time, and their vehicle’s power consumption in real-time. One of many potential benefits is to automatically inform the driver of CO2-reducing behavior by suggesting alternative routes, mode changes using public transportation, substitute destinations, and feasible departure times. Next to end users, traffic control agencies also benefit from projecting more accurate future road states to manage real and hypothetical transportation systems online. The results of the system need to be both faster than real-time and accurate, while the road network under investigation may be large. We propose new route prediction models (along with modified existing models), a method for the efficient superimposed execution of microscopic traffic simulation by using shared state variables, and supervised models to estimate the vehicle-specific power (VSP).